['Tower of London': mental planning, validity and the ceiling effect]

Abstract

Introduction: The Tower of London (TL) is a neuropsychological test that is used to evaluate the capacity for planning. Different versions of the TL exist, including the classical version that presents certain psychometric problems (for example, the ceiling effect).

Aims: The aim of this study was to propose a new method of administering and correcting the classical version of the TL, thus increasing the mental resolution of the planning task through instruction, and varying the system of scoring in order to obtain a wider range of scores.

Patients and methods: Two versions of the TL were administered to two independent, counterbalanced groups. One of them was the one designed by Krikorian and the other was the one proposed for this study. To compare the two versions, Porteus mazes were administered in each group as a criterion variable in the study of correlations.

Results: Both versions correlate with the Porteus mazes in a similar fashion. The variances found in each group are statistically different and the coefficient of covariance is clearly wider in the group that received the version proposed in this study.

Conclusions: Changing the instruction for the mental resolution of planning tasks does not appear to alter the TL. The changes introduced in the correction and the system of scoring of the version proposed for this study present wider variance, which means that the ceiling effect has been overcome.